A common technique to assess wear is to measure loss from a nozzle exposed to propellant combustion gases. Recent experiments in separate laboratories have rated that the scatter from the mass losses recorded for a series of shots with the same propellant is well outside experimental error. This suggests the scatter in data reflects the actual wear process itself and that understanding why the wear process seems erratic should be a key feature of any model of gun barrel wear, if such scatter is also characteristics of wear in guns. To see if the scatter in wear measurements seen with nozzles holds for large caliber guns, a survey was conducted to find wear tests meeting the following criteria: wear vs rounds fired was linear; frequent measurements were made; and the tubes were not chromium-plated. Data from fifteen such tests were collected ranging from 60mm to 155mm cannons. It was shown that the scatter in data in the large caliber guns was the same as seen in the nozzles. Typically, the sample standard deviation was 20-30 percent as large as the sample mean wear/round. Not only does this scatter represent a fruitful area of research for understanding how guns wear, but the inherent scatter must be kept in mind when designing gun wear tests.